Study Of A Multiple Visual Color Sorter Service Platform Based On Internet Of Things

With the improvement of people's attention to food safety and food refinement,the demand for color selectors is on the rise.However,the traditional intelligent control level of color selector is low,which consumes a lot of manpower and material resources in training,maintenance,management and other aspects.In view of this problem,combined with the actual needs of a certain color selector enterprise in Anhui,this paper developed a set of multiple visual color selector service platform based on Internet of Things,which can be applied to industrial PC,web page,mobile terminal and other multiple platforms.The specific research content is as follows:(1)To model the system architecture.The system is divided into three layers: device terminal data acquisition and control layer,server side network communication layer and client side monitoring center layer.The three parts interact with SQL Server database as the core.(2)To design the communication architecture of system.It develops a server monitoring software based on Client/Server architecture for reading database instructions,packaging and sending to color picker,and then receiving and analyzing the data stream and storing it in the database;The system platform of Broswer/Server architecture issues control instructions and visualizes data.A series of optimizations are made to I/O Completion Ports model to create a high performance server communication layer.(3)To develop a human-computer interaction software of upper computer of color sorter.A series of improvements are made to the feature extraction and material sorting algorithm of the classic convolutional neural network model LeNet-5,and the upper computer control program of the color sorting machine based on Windows system is written to replace the configuration screen control system.The bidirectional communication between upper computer and lower computer based on Modbus ASCII is realised.(4)To construct a comprehensive management platform on the Web and APP ends.The platform has seven modules: real-time monitoring,parameter modification,equipment debugging,equipment archives,alarm setting,operation and maintenance management and user management.The visual management function of equipment information is mainly introduced,including measures taken to improve the speed and smoothness of map drawing and point loading.In order to realize the industrialization of the platform,the system was deployed and the performance was tested.The test results showed that the accuracy,reliability and stability of the system color and shape sorting all meet the design requirements,and the cost analysis results showed that it has a certain industrialization value.